0

How Table Load Order and MV Load Order work?

For Incorta 3.x table load order, does higher mean it will load sooner or later. ie does 11 load before 10? also do materialized views and tables both use the same queue, meaning if I set an mv at 11 and a table at 10, which loads first?

3replies Oldest first
  • Oldest first
  • Newest first
  • Active threads
  • Popular
  • You can set the table load order and MV load at the schema level.  Incorta allows you to mix tables and MVs within a single schema.  MV will be executed after all tables are extracted and the related tables (or other MVs) are compacted.  These two setting are independent from each other.  

    The table order with a lower number will be executed earlier than those table with a higher number. Incorta lets you to set multiple tables with the same table load order number.  Incorta will finish all tables with the same table load order number before moving to execute those tables in the next number.

    The MV phase will start after the last table extraction is done and start with those MVs with the lowest number.  

    When a MV is referencing another MV.  In other word, using the data from MV A to build the data in MV B, please set the MV load order to ensure that the latest refreshed data in MV A will be used for refreshing the data in MV B.  

    The schema pool size and table pool size still control the concurrent jobs to be executed.  The tables with the same table load order may not be always extracted concurrently.

    Also, the Table Load Order and MV Load Order only controls the execution during  the extraction and transform (MV)  phases.  Loading data into Incorta engine is not controlled by the Table Load Order or MV Load Order.  The phase of loading the extracted or transformed data into Incorta engine (Loader service in Incorta 4.x) happens after the extraction phase and transform phase are finished.

    Incorta calculate the joins in the Load phase after loading data into engine.  Therefore, the Table Load Order does not really affect the join creation.  However, if you want to guarantee that by the time when the child table (or Fact table) join is being calculated, the parent table (or Dimension) data is already populated, you may want to extract the child table first.  

    Like
  • Hi Dylan,

     

     I have created two MV's. MV 1(CUST_TRX_V) is loaded sucessfully with the data. But MV 2(SALES_BY_CUSTOMER_MV) dependent on MV 1, it is created but there is no data( I mean extracted rows, loaded rows are '0'). Can you plz help me on this.

    I am using Incorta Cloud trail version(4.8.1). Is there any Limitations for this version. Please check the images.

    Is there any size limitation for Incorta cloud trial application.

     

    Thanks

    Anil

    Like
    • Anil Kumar Katari When we have MV to MV dependency within the same schema, as of the release 4.8.1, we will need to define the MV Order by putting them into different group.

      In the below example, EBS_FA_STG_ROWCOUNT MV will be start and be finished for any of the MV in Group 2 starts.

      Like
Like Follow
  • 1 mth agoLast active
  • 3Replies
  • 117Views
  • 3 Following

Product Announcement


We are happy to
announce Incorta 4.8 !!!